Moonshot faces US claims over Kimi training
- US agencies have accused Moonshot AI of extracting capabilities from American AI models to train Kimi.
- The agencies say six Chinese AI firms used billions of tokens across millions of requests since late 2024.
- The allegations come after Kimi K3 triggered market concerns about US leadership in advanced AI.
Moonshot AI has been accused by US agencies of using American AI models to help train its Kimi systems.
The NSA, FBI and CISA said six Chinese AI firms extracted billions of tokens through millions of requests since late 2024.
The agencies said the companies used proxy services and multiple accounts to avoid detection and usage limits.
US agencies have accused Moonshot AI of using Claude and GPT outputs to train Kimi, while the company has yet to comment on the claims.
The agencies described the activity as “malicious knowledge distillation” and said it likely occurred with Chinese government awareness.
The allegations follow Kimi K3’s strong coding performance, which briefly pressured technology shares, Bitcoin (CRYPTO:BTC) and other cryptocurrencies.
The agencies’ claims add a new security concern around Chinese AI development, while Moonshot has yet to publicly respond to the allegations.
